The purpose of the experiments was to analyse the spatial cueing effects of the movements of soccer players executing normal and deceptive (step-over) turns with the ball. Stimuli comprised normal resolution or point-light video clips of soccer players dribbling a football towards the observer then turning right or left with the ball. Clips were curtailed before or on the turn (−160, −80, 0 or +80 ms) to examine the time course of direction prediction and spatial cueing effects. Participants were divided into higher-skilled (HS) and lower-skilled (LS) groups according to soccer experience. In experiment 1, accuracy on full video clips was higher than on point-light but results followed the same overall pattern. Both HS and LS groups correctly identified direction on normal moves at all occlusion levels. For deceptive moves, LS participants were significantly worse than chance and HS participants were somewhat more accurate but nevertheless substantially impaired. In experiment 2, point-light clips were used to cue a lateral target. HS and LS groups showed faster reaction times to targets that were congruent with the direction of normal turns, and to targets incongruent with the direction of deceptive turns. The reversed cueing by deceptive moves coincided with earlier kinematic events than cueing by normal moves. It is concluded that the body kinematics of soccer players generate spatial cueing effects when viewed from an opponent's perspective. This could create a reaction time advantage when anticipating the direction of a normal move. A deceptive move is designed to turn this cueing advantage into a disadvantage. Acting on the basis of advance information, the presence of deceptive moves primes responses in the wrong direction, which may be only partly mitigated by delaying a response until veridical cues emerge.